Are the Days of Undetectable AI Content Coming to an End?

For the past few years, people have been using AI tools to help write emails, blog posts, reports and marketing content, often without anyone knowing. That may be starting to change.

AI company Anthropic is introducing an invisible watermarking system for text generated by its latest Claude models. Unlike traditional watermarks that can be seen on a document or image, this one works behind the scenes. It relies on subtle patterns in the words and phrasing the AI chooses, making it possible to estimate whether Claude was involved in creating the text.

What makes this particularly interesting is that simply copying the content into a new document, email, website or content management system will not automatically remove the watermark. In some cases, even light editing may leave enough of the pattern behind for detection tools to recognise AI involvement.

Before anyone starts worrying, it's important to understand what these tools can and can't do.

They are designed to estimate whether Claude helped generate the text. They cannot identify who used the AI, reveal the prompts that were entered, or determine exactly how much of the final document came from Claude. In fact, a document could potentially carry traces of the watermark even if AI was only used to rewrite or polish a small section.

For businesses, this doesn't mean AI should be avoided. AI can be a valuable tool for drafting content, improving wording, summarising information and boosting productivity.

What it does mean is that organisations should be aware that AI involvement may become more visible over time. Content sent to customers, clients or the public still needs human review to ensure it is accurate, appropriate and reflects the organisation's voice and standards.

AI can help create content faster, but it doesn't remove the responsibility for checking facts, correcting mistakes or making sure messages align with your business values.

As AI technology continues to evolve, transparency around its use is likely to become increasingly important. Whether you're using AI to write a blog post, draft a proposal or improve an email, it's worth considering how your organisation wants to approach AI-generated content and what policies should guide its use.

If your team is already using AI at work and you're not sure what guidelines, policies or safeguards you should have in place, let's have a chat.